The History of the First AI-Generated Image

The Genesis of Synthetic Canvas: The Story of the First AI-Generated Image

Long before neural networks could summon hyper-realistic oil paintings or fantastical landscapes from a simple text prompt, the intersection of art and artificial intelligence began in a quiet laboratory in the late 1960s.

While today’s generative tools feast on billions of parameters and vast internet datasets, the world’s first AI-generated art emerged from a deliberate, painstaking effort to codify human creativity itself.

Meet AARON: The Pioneer AI Artist

The story of the first AI-generated image belongs to Harold Cohen, a renowned British painter and professor. In 1968, at the University of California, San Diego, Cohen made a radical pivot that baffled his contemporaries: he abandoned his physical paintbrush to write code.

Cohen wasn’t trying to build a machine that could copy existing artwork. Instead, he wanted to answer a profound philosophical question: What are the absolute minimum cognitive rules required to make and understand a drawing?

By the early 1970s—with a major milestone reached around 1973—Cohen’s software project, named AARON, produced its first autonomous visual works.

How AARON Made Art Without Deep Learning

AARON was entirely different from modern AI art generators. It did not use machine learning, neural networks, or scraped internet data. Instead, it relied on symbolic, rule-based programming (often referred to in computer science history as GOFAI—Good Old-Fashioned AI).

• Manual Logic: Cohen meticulously translated human artistic principles into code. He taught AARON how to conceptualize basic shapes, handle proportions, overlap objects to create a sense of space, and manage composition.
• Autonomous Decision-Making: Once the rules were set, AARON took the wheel. The program made its own composition choices, deciding where to place lines, how to construct organic figures, and how to balance the elements on the page.
• Physical Execution: Early versions of AARON didn’t just live on a monitor; Cohen hooked the program up to custom-built, turtle-like drawing machines that physically held a pen and plotted the computer’s creations onto paper.

From Primitive Lines to a Legacy

AARON’s earliest outputs were modest—striking, abstract, black-and-white line drawings. Yet, they possessed genuine autonomy. The computer was executing code derived from human logic to create entirely novel compositions that Cohen himself hadn’t pre-drawn or visualized.

As decades passed, Cohen continuously updated AARON. The software evolved from handling simple shapes to populating vast scenes filled with plants, rocks, and human figures, eventually incorporating color and being showcased in prestigious institutions like the Tate Modern and the Whitney Museum of American Art.

Why the First AI Image Still Matters

Looking back from an era dominated by instant text-to-image synthesis, AARON’s primitive lines remind us of a fundamental truth: the desire to co-create with machines is not a modern fad.